Conditionally replicative adenoviruses (CRAds) represent an important novel approach for cancer therapy. CRAd agents are designed to specifically replicate in and kill tumor cells to yield an effective yet safe therapeutic outcome. Although their great potential has awarded them rapid translation into human clinical trials where the safety of CRAds have been clearly highlighted, little data have been obtained with regards to the critical functions of CRAds efficient infection of tumor cells, tumor-specific replication, and lateral spread. Without deeper understanding of the nature of these replicative agents especially in a patient context, further development of CRAds would be greatly hindered. The crux of a general dilemma in the virotherapy field is lack of a monitoring system compatible with replicative agents. Unfortunately, most gene therapy vector detection schemes have been specifically designed to assess gene expression. These modalities are unsuitable for monitoring of CRAds because their very nature is to kill infected tumor cells, a concept at odds with the notion of viable transgene expression. Furthermore, reporters by themselves cannot accurately depict the underlying level of replication as well as true physical distribution of viral progeny, two crucial functions of replicative agents. Clinical trials to date have had to rely on traditional histological analysis of biopsy specimens which are error-prone and cannot portray the multiplicative nature of CRAds. The ideal monitoring system for CRAds should embody the following features: (1) report the level of viral replication, (2) allow direct detection of viral spread, (3) permit dynamic detection of viral activity, (4) minimally perturb replication and spread efficiency of the virus, and (5) possess the power for noninvasive detection. To address this issue, we hypothesized that a genetic adenovirus labeling system using a structural reporter fusion protein would dynamically represent viral replication and spread. We propose adenovirus capsid labeling with pIX-HSV-tkto allow PET-based imaging. In addition, labeling with both EGFP and HSV-tk would allow imaging analysis by both optical and PET-based methods. It is evident that a monitoring system for CRAds is needed for advancing the field. A genetic structural labeling system for adenovirus would offer noninvasive dynamic detection of replication and spread. Not only would this system be indispensable in developing advanced CRAds, it would also be applicable for monitoring CRAd therapy in patients. [unreadable] [unreadable]